Have you heard? My new book Continuous Discovery Habits is now available. Get the product trio's guide to a structured and sustainable approach to continuous discovery. 
Old product backlogs can be the place where ideas go to die. But they can also be a treasure trove of feedback from your customers. And in some cases, they might be a mix of both. The trouble is, you might not know until you spend some time reviewing your specific backlog. And is it even worth taking time to go through this process?
Old product backlogs can be the place where ideas go to die. But they can also be a treasure trove of feedback from your customers. And in some cases, they might be a mix of both. – Tweet This
If you’re a product manager who’s coming into a new team, chances are high that you’re inheriting a backlog. And depending on how long it’s been kept and how carefully it’s been managed, it might be full of hundreds of entries.
So what should you do with them?
Today’s Ask the Community was inspired by a topic that came up in the Continuous Discovery Habits community: What do you do with a giant backlog? The question-asker said that when they searched online outside the community for advice, the general consensus seemed to be to throw old backlogs in the trash. But they wondered whether there might be a way to get some value out of them.
In this post, we’ll share both advice from the community and Teresa’s take on this topic.
Question: I inherited a giant product backlog of feature requests and bugs. What should I do with it? Would you recommend using an opportunity solution tree (or set of opportunity solution trees) to sift through that backlog and try to categorize them?
This question sparked a lively discussion, so here’s a summary of the CDH community discussion as well as Teresa’s thoughts and recommendations.
Raphael Weiner, co-founder of Orbital, shared, “I’d want to first anchor any attempt at making sense of the backlog to the current context. If it’s a backlog full of feature requests and bugs reported directly from customers, it’s already much more valuable than a backlog full of internal ‘ideas.’ I wouldn’t throw it away. But you very well may want to ignore it for the time being if you’re just getting ramped into this new team/context. This might not be your highest value activity right now.”
If your backlog is full of feature requests and bugs reported directly from customers, it’s already much more valuable than a backlog full of internal ‘ideas.’ I wouldn’t throw it away. – Tweet This
He also recommended asking a predecessor, new team members, or stakeholders for their views on it.
He asked if the backlog items already had the contextual data (like customer, revenue, impact, etc.) that someone would need to make sense of them. “Depending on how the feature requests were collected, they may not have the context you’d need to do anything with them other than further investigate anyways,” wrote Raphael.
He also suggested that product backlog items tend to have a short half-life, so once they get too old, they may no longer be as valuable.
Finally, he noted that, “At a minimum, the feature requests are an invitation for a further conversation—i.e. customer interview—and that in itself is a gift!”
Amanda Wernicke, a product manager at Convo, wrote about the importance of defining your outcome first: “Starting with a product outcome will help you to identify whether the items in your giant backlog will help you achieve that outcome or are not relevant.”
Starting with a product outcome will help you to identify whether the items in your giant backlog will help you achieve that outcome or are not relevant. – Tweet This
Lucy Spence, a Senior Technical Product Manager at Octopus Deploy, wrote, “Depending on your comfort of sharing the data, it might actually be one of the situations that using AI to summarize key themes saves you a chunk of time.” Lucy suggested:
- If you use Dovetail, you could load it into Dovetail and ask it to summarize themes. You’ll not get any super useful insights, but it probably will do a decent job of summarizing the high-level themes.
- You could also use ChatGPT, though Lucy wasn’t sure how it would handle being loaded up with significant amounts of data.
- Google will be able to tell you more options. But again, Lucy thought most solutions would provide high-level themes rather than specific insights.
Teresa’s Take: Differentiate Between Using Your Backlog as a Historical Record and a Source of Ideas for What to Build Next
Next, Teresa weighed in with her take. She wrote that she would separate two things:
- Keeping an archive of the backlog around so nothing gets lost. You have a historical record of issues that came up in the past.
- Identifying what to work on next.
“I wouldn’t use your current backlog to decide what to work on next, unless you need to have a few quick wins while you figure out where you can have the most impact,” she continued.
I wouldn’t use your current backlog to decide what to work on next, unless you need to have a few quick wins while you figure out where you can have the most impact. – Tweet This
As for what to do with the historical record, Teresa recommended referencing it throughout discovery. For example, once you set an outcome, Teresa suggested looking to see what has come up in the past that might be related to that outcome. Once you choose a target opportunity, you can search to see what might be related to that.
Finally, Teresa wrote, “If you are allowed to use generative AI at work, something like ChatGPT or Claude might be an effective way to understand what’s in there.”
She said she’d recommend using a Teams or Enterprise account so they don’t use your data for training purposes. You also might want to take a pass through it to make sure there is no personally identifying information in the data set.
Depending on what tool you are using for your backlog, you should be able to export everything into CSV, Excel, or something similar. You can then upload that to ChatGPT and ask it questions.
But it’s worth noting that you may have to iterate on your prompt. Teresa recommended starting with something like this:
You are a product manager in a new role and inherited the following backlog of items. Think through how to make sense of it. Here are some step-by-step instructions to apply:
- What are the most common themes or topics in the list?
- How have the most common themes or topics changed over time?
- Are there any duplicates in the list? If two items are semantically similar, consider them a duplicate. For any duplicates, list the item and how often it comes up.
“If you haven’t used ChatGPT before, think about it like having a conversation with an intern,” wrote Teresa. “You don’t always get what you ask for the first time. But it takes feedback well. So if the first response isn’t very good, give it feedback and tell it to try again.”
If you haven’t used ChatGPT before, think about it like having a conversation with an intern. You don’t always get what you ask for the first time. But it takes feedback well. – Tweet This
Later as you start to identify outcomes and opportunities, you can ask ChatGPT if there are any related backlog items to that outcome or that opportunity.
We regularly tackle questions like this in the Continuous Discovery Habits Community. If you’re looking for a safe space where you can share your challenges and get feedback from like-minded peers, come join us there!
